mirror of https://github.com/python/cpython
302 lines
8.6 KiB
Python
302 lines
8.6 KiB
Python
# Check every path through every method of UserDict
|
|
|
|
import unittest
|
|
from test import test_support, mapping_tests
|
|
import UserDict
|
|
|
|
d0 = {}
|
|
d1 = {"one": 1}
|
|
d2 = {"one": 1, "two": 2}
|
|
d3 = {"one": 1, "two": 3, "three": 5}
|
|
d4 = {"one": None, "two": None}
|
|
d5 = {"one": 1, "two": 1}
|
|
|
|
class UserDictTest(mapping_tests.TestHashMappingProtocol):
|
|
type2test = UserDict.IterableUserDict
|
|
|
|
def test_all(self):
|
|
# Test constructors
|
|
u = UserDict.UserDict()
|
|
u0 = UserDict.UserDict(d0)
|
|
u1 = UserDict.UserDict(d1)
|
|
u2 = UserDict.IterableUserDict(d2)
|
|
|
|
uu = UserDict.UserDict(u)
|
|
uu0 = UserDict.UserDict(u0)
|
|
uu1 = UserDict.UserDict(u1)
|
|
uu2 = UserDict.UserDict(u2)
|
|
|
|
# keyword arg constructor
|
|
self.assertEqual(UserDict.UserDict(one=1, two=2), d2)
|
|
# item sequence constructor
|
|
self.assertEqual(UserDict.UserDict([('one',1), ('two',2)]), d2)
|
|
self.assertEqual(UserDict.UserDict(dict=[('one',1), ('two',2)]), d2)
|
|
# both together
|
|
self.assertEqual(UserDict.UserDict([('one',1), ('two',2)], two=3, three=5), d3)
|
|
|
|
# alternate constructor
|
|
self.assertEqual(UserDict.UserDict.fromkeys('one two'.split()), d4)
|
|
self.assertEqual(UserDict.UserDict().fromkeys('one two'.split()), d4)
|
|
self.assertEqual(UserDict.UserDict.fromkeys('one two'.split(), 1), d5)
|
|
self.assertEqual(UserDict.UserDict().fromkeys('one two'.split(), 1), d5)
|
|
self.assert_(u1.fromkeys('one two'.split()) is not u1)
|
|
self.assert_(isinstance(u1.fromkeys('one two'.split()), UserDict.UserDict))
|
|
self.assert_(isinstance(u2.fromkeys('one two'.split()), UserDict.IterableUserDict))
|
|
|
|
# Test __repr__
|
|
self.assertEqual(str(u0), str(d0))
|
|
self.assertEqual(repr(u1), repr(d1))
|
|
self.assertEqual(`u2`, `d2`)
|
|
|
|
# Test __cmp__ and __len__
|
|
all = [d0, d1, d2, u, u0, u1, u2, uu, uu0, uu1, uu2]
|
|
for a in all:
|
|
for b in all:
|
|
self.assertEqual(cmp(a, b), cmp(len(a), len(b)))
|
|
|
|
# Test __getitem__
|
|
self.assertEqual(u2["one"], 1)
|
|
self.assertRaises(KeyError, u1.__getitem__, "two")
|
|
|
|
# Test __setitem__
|
|
u3 = UserDict.UserDict(u2)
|
|
u3["two"] = 2
|
|
u3["three"] = 3
|
|
|
|
# Test __delitem__
|
|
del u3["three"]
|
|
self.assertRaises(KeyError, u3.__delitem__, "three")
|
|
|
|
# Test clear
|
|
u3.clear()
|
|
self.assertEqual(u3, {})
|
|
|
|
# Test copy()
|
|
u2a = u2.copy()
|
|
self.assertEqual(u2a, u2)
|
|
u2b = UserDict.UserDict(x=42, y=23)
|
|
u2c = u2b.copy() # making a copy of a UserDict is special cased
|
|
self.assertEqual(u2b, u2c)
|
|
|
|
class MyUserDict(UserDict.UserDict):
|
|
def display(self): print self
|
|
|
|
m2 = MyUserDict(u2)
|
|
m2a = m2.copy()
|
|
self.assertEqual(m2a, m2)
|
|
|
|
# SF bug #476616 -- copy() of UserDict subclass shared data
|
|
m2['foo'] = 'bar'
|
|
self.assertNotEqual(m2a, m2)
|
|
|
|
# Test keys, items, values
|
|
self.assertEqual(u2.keys(), d2.keys())
|
|
self.assertEqual(u2.items(), d2.items())
|
|
self.assertEqual(u2.values(), d2.values())
|
|
|
|
# Test has_key and "in".
|
|
for i in u2.keys():
|
|
self.assert_(u2.has_key(i))
|
|
self.assert_(i in u2)
|
|
self.assertEqual(u1.has_key(i), d1.has_key(i))
|
|
self.assertEqual(i in u1, i in d1)
|
|
self.assertEqual(u0.has_key(i), d0.has_key(i))
|
|
self.assertEqual(i in u0, i in d0)
|
|
|
|
# Test update
|
|
t = UserDict.UserDict()
|
|
t.update(u2)
|
|
self.assertEqual(t, u2)
|
|
class Items:
|
|
def items(self):
|
|
return (("x", 42), ("y", 23))
|
|
t = UserDict.UserDict()
|
|
t.update(Items())
|
|
self.assertEqual(t, {"x": 42, "y": 23})
|
|
|
|
# Test get
|
|
for i in u2.keys():
|
|
self.assertEqual(u2.get(i), u2[i])
|
|
self.assertEqual(u1.get(i), d1.get(i))
|
|
self.assertEqual(u0.get(i), d0.get(i))
|
|
|
|
# Test "in" iteration.
|
|
for i in xrange(20):
|
|
u2[i] = str(i)
|
|
ikeys = []
|
|
for k in u2:
|
|
ikeys.append(k)
|
|
keys = u2.keys()
|
|
self.assertEqual(set(ikeys), set(keys))
|
|
|
|
# Test setdefault
|
|
t = UserDict.UserDict()
|
|
self.assertEqual(t.setdefault("x", 42), 42)
|
|
self.assert_(t.has_key("x"))
|
|
self.assertEqual(t.setdefault("x", 23), 42)
|
|
|
|
# Test pop
|
|
t = UserDict.UserDict(x=42)
|
|
self.assertEqual(t.pop("x"), 42)
|
|
self.assertRaises(KeyError, t.pop, "x")
|
|
self.assertEqual(t.pop("x", 1), 1)
|
|
t["x"] = 42
|
|
self.assertEqual(t.pop("x", 1), 42)
|
|
|
|
# Test popitem
|
|
t = UserDict.UserDict(x=42)
|
|
self.assertEqual(t.popitem(), ("x", 42))
|
|
self.assertRaises(KeyError, t.popitem)
|
|
|
|
##########################
|
|
# Test Dict Mixin
|
|
|
|
class SeqDict(UserDict.DictMixin):
|
|
"""Dictionary lookalike implemented with lists.
|
|
|
|
Used to test and demonstrate DictMixin
|
|
"""
|
|
def __init__(self, other=None, **kwargs):
|
|
self.keylist = []
|
|
self.valuelist = []
|
|
if other is not None:
|
|
for (key, value) in other:
|
|
self[key] = value
|
|
for (key, value) in kwargs.iteritems():
|
|
self[key] = value
|
|
def __getitem__(self, key):
|
|
try:
|
|
i = self.keylist.index(key)
|
|
except ValueError:
|
|
raise KeyError
|
|
return self.valuelist[i]
|
|
def __setitem__(self, key, value):
|
|
try:
|
|
i = self.keylist.index(key)
|
|
self.valuelist[i] = value
|
|
except ValueError:
|
|
self.keylist.append(key)
|
|
self.valuelist.append(value)
|
|
def __delitem__(self, key):
|
|
try:
|
|
i = self.keylist.index(key)
|
|
except ValueError:
|
|
raise KeyError
|
|
self.keylist.pop(i)
|
|
self.valuelist.pop(i)
|
|
def keys(self):
|
|
return list(self.keylist)
|
|
def copy(self):
|
|
d = self.__class__()
|
|
for key, value in self.iteritems():
|
|
d[key] = value
|
|
return d
|
|
def fromkeys(cls, keys, value=None):
|
|
d = cls()
|
|
for key in keys:
|
|
d[key] = value
|
|
return d
|
|
fromkeys = classmethod(fromkeys)
|
|
|
|
class UserDictMixinTest(mapping_tests.TestMappingProtocol):
|
|
type2test = SeqDict
|
|
|
|
def test_all(self):
|
|
## Setup test and verify working of the test class
|
|
|
|
# check init
|
|
s = SeqDict()
|
|
|
|
# exercise setitem
|
|
s[10] = 'ten'
|
|
s[20] = 'twenty'
|
|
s[30] = 'thirty'
|
|
|
|
# exercise delitem
|
|
del s[20]
|
|
# check getitem and setitem
|
|
self.assertEqual(s[10], 'ten')
|
|
# check keys() and delitem
|
|
self.assertEqual(s.keys(), [10, 30])
|
|
|
|
## Now, test the DictMixin methods one by one
|
|
# has_key
|
|
self.assert_(s.has_key(10))
|
|
self.assert_(not s.has_key(20))
|
|
|
|
# __contains__
|
|
self.assert_(10 in s)
|
|
self.assert_(20 not in s)
|
|
|
|
# __iter__
|
|
self.assertEqual([k for k in s], [10, 30])
|
|
|
|
# __len__
|
|
self.assertEqual(len(s), 2)
|
|
|
|
# iteritems
|
|
self.assertEqual(list(s.iteritems()), [(10,'ten'), (30, 'thirty')])
|
|
|
|
# iterkeys
|
|
self.assertEqual(list(s.iterkeys()), [10, 30])
|
|
|
|
# itervalues
|
|
self.assertEqual(list(s.itervalues()), ['ten', 'thirty'])
|
|
|
|
# values
|
|
self.assertEqual(s.values(), ['ten', 'thirty'])
|
|
|
|
# items
|
|
self.assertEqual(s.items(), [(10,'ten'), (30, 'thirty')])
|
|
|
|
# get
|
|
self.assertEqual(s.get(10), 'ten')
|
|
self.assertEqual(s.get(15,'fifteen'), 'fifteen')
|
|
self.assertEqual(s.get(15), None)
|
|
|
|
# setdefault
|
|
self.assertEqual(s.setdefault(40, 'forty'), 'forty')
|
|
self.assertEqual(s.setdefault(10, 'null'), 'ten')
|
|
del s[40]
|
|
|
|
# pop
|
|
self.assertEqual(s.pop(10), 'ten')
|
|
self.assert_(10 not in s)
|
|
s[10] = 'ten'
|
|
self.assertEqual(s.pop("x", 1), 1)
|
|
s["x"] = 42
|
|
self.assertEqual(s.pop("x", 1), 42)
|
|
|
|
# popitem
|
|
k, v = s.popitem()
|
|
self.assert_(k not in s)
|
|
s[k] = v
|
|
|
|
# clear
|
|
s.clear()
|
|
self.assertEqual(len(s), 0)
|
|
|
|
# empty popitem
|
|
self.assertRaises(KeyError, s.popitem)
|
|
|
|
# update
|
|
s.update({10: 'ten', 20:'twenty'})
|
|
self.assertEqual(s[10], 'ten')
|
|
self.assertEqual(s[20], 'twenty')
|
|
|
|
# cmp
|
|
self.assertEqual(s, {10: 'ten', 20:'twenty'})
|
|
t = SeqDict()
|
|
t[20] = 'twenty'
|
|
t[10] = 'ten'
|
|
self.assertEqual(s, t)
|
|
|
|
def test_main():
|
|
test_support.run_unittest(
|
|
UserDictTest,
|
|
UserDictMixinTest
|
|
)
|
|
|
|
if __name__ == "__main__":
|
|
test_main()
|